Download Sharing clinical genomic data for better diagnostics

BRIEFING NOTE | Delivering clinical whole genome analysis
October 2014
Authors
Dr Sobia Raza
Dr Leila Luheshi
Alison Hall
Contact
Dr Sobia Raza
[email protected]
Sharing clinical
genomic data for
better diagnostics
Accurate determination of the genetic cause of a rare
disease depends on the availability of knowledge
about which mutations in a patient’s genome could
underlie their condition. As genomic research and
analysis expands, this knowledge base is increasing
rapidly but access to it, even within the National
Health Service, is often severely restricted by inability
or unwillingness to share this data across and even
within institutions. This situation is incompatible with
the delivery of high-quality, efficient, and reliable,
genomic medicine in the NHS. The barriers to
sharing genomic and clinical data must therefore be
recognised and addressed as a matter of urgency.
Well executed whole genome analysis offers an unprecedented opportunity
to determine the genetic basis of thousands of undiagnosed rare diseases, but
a key bottleneck in delivery is the ongoing struggle to analyse and interpret
complex and voluminous genome sequence data. This necessitates the
comparison of a patient genome with current genomic, genetic and clinical
data on the wider population and other patients with the same or similar
illnesses. It is therefore essential for clinicians and scientists to have routine
access to all existing information on the genetic variants and associated clinical
characteristics relevant to their patient’s disorder, in order to make the most
informed interpretation and diagnosis.
This briefing explores the need for sharing genomic data within the NHS
for delivering clinical genetic diagnoses, and sets out recommendations for
addressing current barriers. Not covered here are the wider socio-political,
ethical and regulatory concerns surrounding genomic and clinical data
sharing within and beyond the NHS, which must also be addressed in order to
balance safeguarding patient interests with advancing genomic medicine.
Page 1 | Sharing clinical genomic data for better diagnostics
BRIEFING NOTE | Delivering clinical whole genome analysis
Current guidance on
genomic data sharing:
Best practice
guidelines followed
by clinically accredited
laboratories in the UK
performing genetics
tests recommend
the submission of
clinically relevant
variants found by the
lab to appropriate
databases.1-2
The guidelines also note
that ‘complete upload
of all variants and
associated phenotype
information from
every patient is the
ultimate goal’, ideally
via automated export
functions.2
What are the barriers to sharing data within the NHS?
Despite recommendations within best practice guidelines [left] and efforts by
a number of international initiatives advocating and encouraging sharing,3-7
knowledge of genomic variants tends to be isolated in silos and data
exchange between laboratories and deposition to existing public databases
(e.g. DECIPHER, ClinVar, LoVD) is ineffective. The reasons for this situation include:
•
No prior urgency to share until now
Traditionally individual genetics laboratories have acquired specialist
knowledge in distinct and non-overlapping disease domains and have
not needed to share their findings more widely, since they are largely the
sole consumers of that information in the UK.
•
Disincentives
Currently genetic testing services in the UK are in competition with
each other to attract NHS referrals for testing and income. Knowledge
generated by a given laboratory, might then be viewed as their
intellectual property, which if distributed more widely would devalue the
exclusivity of their service.
•
Overheads and absence of a sustainable solution
The logistical barriers to sharing data (time, resource overheads, and
insufficient infrastructure) can, for many laboratories, outweigh the
benefits reaped, particularly in the absence of a sustainable strategy for
aggregating data from all diagnostic molecular genetics laboratories.
•
Social, regulatory and privacy concerns.
There are multiple and widespread social, legal and ethical concerns
around the sharing of patients’ personal genomic and clinical information.
Sharing genomic data – why is action needed now?
The promise of genomic medicine can only be realised by aggregating
genomic data and knowledge. By enabling better data sharing now, the NHS
can accomplish efficiency gains, patient benefits, and meet the target of
becoming world leaders in genomic medicine.
The benefits to the NHS and patients of sharing genomic data:
•
More, faster, and more accurate diagnoses
Data sharing will increase the number of identified disease causing
variants and knowledge of disease associated genes.
•
Improved equality of access
Better data sharing will improve the quality and efficiency of diagnosis
across all NHS genetic testing facilities.
•
Benefits and savings across multiple NHS specialities
Patients with genetic disorders are also assessed in other clinical
departments (e.g. cardiology, paediatrics, ophthalmology).
Page 2 | Sharing clinical genomic data for better diagnostics
•
•
Better and more personalised treatment
More genomic data will better serve an ethnically diverse UK population,
by improving knowledge on genetic diversity across different ethnic
groups.
New medical and public health interventions
Underpinned by greater understanding of the relationship between
genes and health.
The risks of not acting now:
•
‘Diagnostic lottery’
Patients fail to receive diagnoses when their sample is referred to a
laboratory that does not have access to information that could lead to
their diagnosis.
•
Misdiagnosis
Leading to inappropriate treatment and care, because the clinician was
unaware of critical information.
•
Missed resolution of cases
Where variants of unknown significance are not reclassified due to lack of
information consolidation – prolonging a patient’s wait for a diagnosis.
•
Inefficient use of resources
Resulting from multiple testing centres trying to reproduce knowledge
that already exists within the NHS.
•
Opportunity cost
Failure to act now may have long-reaching effects.
What data needs to be shared?
All types of variants
Aggregating information on patient variants and clinical data is vital to
discover significant links between specific variants and disease.
Sequence data
Essential to determine what ‘normal’ variation is within populations, and
use this information to exclude the vast number of variants unlikely to be
responsible for disease and improve the chances of finding ‘true’ variants
in a patient’s genome.
Phenotype and clinical information
Understanding the clinical context of a previously identified mutation and
accessing detailed information on patients’ disease characteristics and
symptoms (phenotype) can help determine whether it is applicable to
another case.
“Health and social
care professionals
should have the
confidence to share
information in the
best interests of
their patients...They
should be supported
by the policies of
their employers,
regulators and
professional bodies”
Information governance
review (2013)8
By enabling better
data sharing
now, the NHS
can accomplish
efficiency gains,
patient benefits, and
meet the target of
becoming world
leaders in genomic
medicine.
Sharing clinical genomic data for better diagnostics | Page 3
BRIEFING NOTE | Delivering clinical whole genome analysis
Better data sharing for improved diagnosis: what next?
Summary
Widespread
sharing of genetic
and clinical data
within the NHS is
necessary to fully
realise the benefits
of genomic
medicine.
Failure to take
action now will
have undesirable
consequences for
patients and the
NHS. Conversely, a
timely solution will
have far-reaching
benefits, and help
position the UK
at the forefront of
genomic medicine.
•
The existing incentive structure undermines the objective of open
data for wider public benefit. To gain buy-in and cooperation from
all stakeholders, a system is needed where all NHS-funded genetic
testing facilities are mandated to share their data with other (approved)
laboratories.
•
A NHS data resource for curating and sharing genomic data is an essential
step towards ensuring long term sustainability and safeguarding patients’
interests and trust.
•
A data resource must serve the needs of clinicians to deliver better
diagnosis, whilst being simple to use and to contribute to. Stakeholders
must therefore work together to establish an effective solution.
•
An NHS genomics repository should have the capacity to interact with
the infrastructure created by the 100,000 Genomes Project. Consideration
should be given to whether consent for this data sharing might be sought
upfront, to ensure timely access to data; transparency with patients about
the use of their genomic data is imperative.
•
In order to affect change and attain widespread acceptability of the
collection of patients clinical and genomic information, data privacy,
confidentiality and ownership concerns must be managed appropriately.
•
Leadership and strategic oversight are required for the successful,
coordinated delivery of a data resource and compliance with data
sharing, and these must therefore be established.
•
Delivery of an improved sharing infrastructure will require investment,
yet will generate cost savings and substantial benefit to patients and the
NHS in the long run. NHS providers and commissioners must cooperate to
devise a feasible, fair and sustainable funding mechanism.
References
Wallis Y et al., ACGS (2013); 2 Ellard S et al., ACGS (2014); 3 Clinvitae database: http://clinvitae.invitae.
com/ ; 4 The Global Alliance for Genomics and Health http://genomicsandhealth.org/ ; 5 Human
Variome Project: www.humanvariomeproject.org/ ; 6 International Collaboration for Clinical Genomics
www.iccg.org/about-the-iccg/clingen/ ; 7 Free the data www.free-the-data.org; 8 The information
governance review (2013).
1
Page 4 | Sharing clinical genomic data for better diagnostics
For more information, visit:
www.phgfoundation.org